Severe destructive behavior represents one of the most significant barriers to community integration for approximately 160,000 persons with developmental disabilities in the U.S. It is the most common reason for an individual with developmental disabilities remaining in or being returned to a state institution. The annual cost of these problems exceeds 3 billion dollars. Our promising approach for improving our understanding of and ability to treat destructive behavior is called functional analysis. The goals of functional analysis are to identify, for a given individual, (a) the environmental contexts (i.e., antecedent conditions) in which aberrant behavior is likely and unlikely to occur, (b) the consequences that reinforce or maintain the behavior, and (c) effective treatments. During a functional analysis, variables hypothesized to affect destructive behavior are systematically manipulated using single-case designs and their effects are directly observed and measured. Recent functional analysis research has conceptualized destructive behavior in terms of choice responding. That is, individuals sometimes choose to display destructive rather than appropriate behavior because the former response class is either less effortful or it produces a better outcome (i.e., produces an increased rate, magnitude, quality, or immediacy of reinforcement). Consistent with this view, providing individuals with enhanced choice-making opportunities via appropriate behavior often decreases problem behavior. The project is designed to extend this line of research by examining (a) the role played by variables that affect choice responding (e.g., reinforcement rate and quality) in the maintenance of destructive behavior and (b) the extent to which these variables may be used to increase treatment effectiveness.